CSV files in Cloud Storage contain mixed data types and inconsistent formats (phone numbers, addresses). Which pipeline approach ensures data quality by cleansing and transforming before loading into BigQuery?

Data Fusion is a fully managed, cloud-native data integration service built on Cloud Data Loss Prevention (DLP) and Cloud Dataflow. It provides a visual, code-free interface for building and managing ETL/ELT pipelines, making it ideal for complex data cleansing and transformation tasks involving mixed data types and inconsistent formats before loading into BigQuery. Using Data Fusion to convert to AVRO first is an extra, unnecessary step; Data Fusion can directly transform CSVs. Loading into a staging table and transforming with SQL is possible but less efficient for complex cleansing than Data Fusion's specialized tools. Transforming in place after loading into a table with the desired schema would fail due to the inconsistent formats and mixed data types in the source CSVs.